1.Assume the following: Ho: P1 - P2 = 0; Ha: P1- P2 > 0; large random samples, alpha = .05. What critical value(s) would lead you to reject the null hypothesis in favor of the alternative hypothesis?
2. Assume the following: Ho: U1 - U2 = 0; Ha: U1- U2 ne 0; small random samples (n1 = 15, n2 = 13); alpha = .05. What critical value(s) would lead you to reject the null hypothesis in favor of the alternative hypothesis?
3. Assume the following: Ho: U1 - U2 = 0; Ha: U1- U2 > 0; small random samples (n1 = 15, n2 = 13); alpha = .05. What critical value(s) would lead you to reject the null hypothesis in favor of the alternative hypothesis?
4. Assume the following: Ho: P1 - P2 = 0; Ha: P1- P2 ne 0; large random samples, alpha = .05. What critical value(s) would lead you to reject the null hypothesis in favor of the alternative hypothesis?
I am desperate for clarity..Thank you in advance
2007-05-30 14:59:01 · 2 answers · asked by tooknowhim 1 in Science & Mathematics ➔ Mathematics
I takes a whole semester of statistics to get to the point where you would be asked such a question. They are asking you to test the null hypothesis using the F-statistic. The procedure should be in your book. Summary is at http://en.wikipedia.org/wiki/F-test. The problem is formulated in a way that you are actually going to disprove something basic, like U1-U2=0. If U1-U2 is not zero, then there is some non-trivial correlation between the functions/vectors/measurements U1 and U2. If you plot points on a graph and you get a straight line up, or across, then one system (e.g., U1) has no effect on the other (U2). The F-test and null hypothesis testing in general allows you to compare data/functions/systems etc. that have more than two parameters. However, at this level they will confine themselves to problems with only two parameters, U1 and U2 for instance. The procedure is the same, but more tedious when there are many sets to compare.
2007-05-30 15:12:45 · answer #1 · answered by sefarkas 2 · 0⤊ 0⤋
The basic difference is that with large random samples, you would use the characteristics of the normal distribution. So in (1), if you want to reject the null hypothesis, you would have to show that P1-P2/ (standard deviation) > +2
In (2) and (3), you use the characteristics of the t-distribution, since you have finite samples.
2007-05-30 15:09:05 · answer #2 · answered by cattbarf 7 · 0⤊ 0⤋